English(EN) Which Negatives Matter? Ask Your Text Encoder: Adaptive Similarity Margins for Dense-Caption Retrieval

新的 HN-CLIP 方法提高了密集字幕检索的准确性和训练速度

作者 PulseAugur 编辑部 · [1 个来源] · 2026-08-20 04:00

研究人员开发了 HN-CLIP，这是一种通过解决标准 InfoNCE 目标中的局限性来改进密集字幕检索的新方法。这种新方法使用文本编码器自身的几何结构为负样本构建自适应相似度边距，为更相似的字幕分配更大的边距，而无需额外数据或复杂处理。在四个基准上的实验表明，与现有的最先进方法相比，HN-CLIP 显著提高了检索准确性并加速了训练。 AI

影响这项研究可能带来更高效、更准确的图像-文本匹配系统，从而惠及图像搜索和内容审核等应用。

排序理由关于密集字幕检索新方法的学术论文。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.AI 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。我们如何撰写摘要 →

本文如何被排名

Signal score: 0 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 关于密集字幕检索新方法的学术论文。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration: Single-source cluster; Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics: paper, model release; Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance: High; Clearly on-topic for AI-industry coverage.
Story freshness: 49 days old; Aged out of breaking-news scoring windows; ranking reflects the durable signal from the full source set.

完整方法见我们的编辑标准。

报道来源 [1]

arXiv cs.AI TIER_1 English(EN) · Haoyue Liu, Ye Chen, Zhichao Wang, Xiaoying Tang · 2026-08-20 04:00

哪些负样本重要？让文本编码器来回答：用于密集字幕检索的自适应相似度边界

arXiv:2608.18521v1 Announce Type: new Abstract: Dense-caption retrieval has recently been improved by introducing segmentation, edge maps, LLM-filtered captions, and cross-modal modules into contrastive fine-tuning. However, these methods largely inherit the same InfoNCE objectiv…

本文如何被排名

报道来源 [1]

哪些负样本重要？让文本编码器来回答：用于密集字幕检索的自适应相似度边界

相关实体

相关话题